Skip to main content

This is currently only available through our private beta

Overview

The Learner Variability Navigator dataset (LVN) helps scale the impact of learning science by making connections between student context and learning goals to recommend instructional practices likely to help a given student reach a specific goal. We’ve included 3 core concepts from the Learner Variability Tool from Digital Promise.

Learner models

Learner models are grade- and subject-specific models (e.g., math for grades 3–6 or literacy for grades 4–6). Each learner model connects student context and learning goals with instructional strategies. They help identify what supports student success for specific subjects and grade bands.

Factors

These are research-backed elements that influence learning across different learner models. These factors include:
  • Learner background (including physical well-being)
  • Self-management and interpersonal skills (such as self-regulation)
  • Cognitive processes (like long-term memory)
  • Subject-specific skills (like geometric reasoning)

Strategies

Strategies are instructional practices that support learning across these factors. Each strategy is supported by research and applies to multiple factors and learner models. For example, the peer teaching strategy supports factors like math communication and long-term memory. 

Data relationship diagram

LVN Data Model Diagram High Level

Entities

LearnerModel

Learner models are pairs of grades and academic subjects (e.g., math for grades 3-6, literacy for grades 4-6). For each learner model, there’s a subset of relevant factors and strategies. As such, they help contextualize what’s important for student success tied to specific grade bands and subjects. We have the following LearnerModels in the Knowledge Graph:
  • Math PK-2
  • Math 3-6
  • Math 7-10
  • Literacy PK-3
  • Literacy 4-6
  • Literacy 7-12
  • Portrait of a Learner PK-3
  • Portrait of a Learner 4-8
  • Portrait of a Learner 9-12

Properties

 These entities have the following metadata:
PropertyDescriptionTypeCardinality
academicSubjectThe academic subjectAcademicSubjectENUM1
attributionStatementA textual credit that acknowledges the source or creator of a work, included when required by the content’s license (e.g., Creative Commons BY). The statement specifies how the creator should be attributed in accordance with the license terms.String1
authorThe author of this content.String1
contentThe materials that make up a learner model along with additional information (if available).String0..1
descriptionA description of the item.String0..1
gradeLevelSpecifies the educational grade for which the curriculum component is intendedArray of GradeLevelENUM0..n
identifierThe identifier of the item, either as textual strings or as URL (URI) links.String1
licenseA license document that applies to this content, typically indicated by URL.String1
nameThe name of the item.String0..1
providerThe service provider, service operator, or service performerString1

Factor

A factor is a characteristic that influences learning or performance. Factors span two categories: Context (e.g., background, environment, access to resources, attention, working memory) and Skills (e.g., decoding, fluency, algebraic reasoning). Factors may connect to other factors and can be linked to instructional strategies.

Properties

 These entities have the following metadata:
PropertyDescriptionTypeCardinality
academicSubjectThe academic subjectAcademicSubjectENUM1
attributionStatementA textual credit that acknowledges the source or creator of a work, included when required by the content’s license (e.g., Creative Commons BY). The statement specifies how the creator should be attributed in accordance with the license terms.String1
authorThe author of this content.String1
categoryA grouping for factors within a LearnerModel.String0..1
citationsReferences to research that supports the factor.Array0..n
contentThe materials that make up a factor along with additional information (if available).String0..1
descriptionA description of the item.String0..1
gradeLevelSpecifies the educational grade for which the curriculum component is intendedArray of GradeLevelENUM0..n
identifierThe identifier of the item, either as textual strings or as URL (URI) links.String1
licenseA license document that applies to this content, typically indicated by URL.String1
nameThe name of the item.String0..1
providerThe service provider, service operator, or service performerString1

Strategy

The specific instructional best practices that support learning across the whole child framework. Each of the research-backed approaches supports a variety of factors and learner models (e.g., the ‘peer teaching’ strategy supports factors like math communication and long-term memory).

Properties

 These entities have the following metadata:
PropertyDescriptionTypeCardinality
academicSubjectThe academic subjectAcademicSubjectENUM1
attributionStatementA textual credit that acknowledges the source or creator of a work, included when required by the content’s license (e.g., Creative Commons BY). The statement specifies how the creator should be attributed in accordance with the license terms.String1
authorThe author of this content.String1
categoryA grouping for strategies within a LearnerModel.String0..1
citationsReferences to research that supports the strategy.Array0..n
contentThe materials that make up a strategy along with additional information (if available).String0..1
descriptionA description of the item.String0..1
gradeLevelSpecifies the educational grade for which the curriculum component is intendedArray of GradeLevelENUM0..n
identifierThe identifier of the item, either as textual strings or as URL (URI) links.String1
licenseA license document that applies to this content, typically indicated by URL.String1
nameThe name of the itemString0..1
providerThe service provider, service operator, or service performerString1

Relationships

The relationships relevant for LVN data are:
  • hasFactor
  • hasStrategy
  • interactsWithFactor
  • targetsFactor
  • relevantToStandard
A hasFactor relationship connects a LearnerModel to one or more Factors included in that model’s profile; one-to-many (:LearnerModel)-[:hasFactor]->(:Factor) A hasStrategy relationship connects a LearnerModel to one or more Strategies recommended or available within that model; one-to-many (:LearnerModel)-[:hasStrategy]->(:Strategy) An interactsWithFactor relationship is a bidirectional association between two Factors representing mutual interplay. (:Factor)-[:interactsWithFactor]->(:Factor) A targetsFactor relationship indicates that a strategy is intended to build or directly influence a Factor; many-to-many (:Strategy)-[:targetsFactor]-(:Factor) A relevantToStandard relationship links a Factor to a StandardFrameworkItem to indicate relevance to the expectations expressed in that StandardFrameworkItem; many-to-many. Note: this relationship only exists for Common Core Math standards for v1.2.0 (:Factor)-[:relevantToStandard]->(:StandardFrameworkItem)